Frequency-driven Imperceptible Adversarial Attack on Semantic Similarity
This addresses the vulnerability of classifiers to imperceptible attacks that generalize across datasets, which is an incremental improvement over existing methods.
The paper tackles the problem of adversarial attacks that lack cross-dataset generalization and are perceptible to humans by proposing a method that attacks semantic similarity on feature representations with low-frequency constraints, resulting in more imperceptible and transferable adversarial examples across datasets and architectures.
Current adversarial attack research reveals the vulnerability of learning-based classifiers against carefully crafted perturbations. However, most existing attack methods have inherent limitations in cross-dataset generalization as they rely on a classification layer with a closed set of categories. Furthermore, the perturbations generated by these methods may appear in regions easily perceptible to the human visual system (HVS). To circumvent the former problem, we propose a novel algorithm that attacks semantic similarity on feature representations. In this way, we are able to fool classifiers without limiting attacks to a specific dataset. For imperceptibility, we introduce the low-frequency constraint to limit perturbations within high-frequency components, ensuring perceptual similarity between adversarial examples and originals. Extensive experiments on three datasets (CIFAR-10, CIFAR-100, and ImageNet-1K) and three public online platforms indicate that our attack can yield misleading and transferable adversarial examples across architectures and datasets. Additionally, visualization results and quantitative performance (in terms of four different metrics) show that the proposed algorithm generates more imperceptible perturbations than the state-of-the-art methods. Code is made available at.